It's a big ocean out there.[/caption] In the early-morning hours of July 24, a Montauk fisherman named John Aldridge fell off a lobster boat in the Atlantic Ocean 40 miles off Long Island, the victim of an onboard mishap with a hook. The boat motored on, its crewmembers unaware for several hours that one of their own had tumbled overboard. As Aldridge floated in the water, wondering whether he would survive the day, his shipmates radioed the U.S. Coast Guard to ask for assistance. According to The New York Times, which documented the rescue operation, the Coast Guard deployed sizable resources to find him—including a sophisticated set of data-analytics tools. Those tools include a simulation that incorporates various data-points such as a person’s weight and height, ocean temperature, and the weather to plot out the chances of survival after plunging into the water; the simulation told the Coast Guard that Aldridge would last a maximum of 19 hours before a combination of hypothermia and exhaustion sent him under. The Coast Guard also relies on software to generate a search grid. “At its heart is a Monte Carlo-style simulator that can generate, in just a few minutes, as many as 10,000 points to represent how far and in what direction a ‘search object’ might have drifted,” is how the Times describes it. “Operators input a variety of data, from the last known location of a lost mariner to the ocean currents and wind direction.” That software’s been in place for seven years, although the Coast Guard has apparently relied on simulations since the 1970s. Algorithms allow the agency to radically reduce a search area, although the output is only as good as the input; Aldridge fell in the water at 3:30 A.M., but the system initially had him in the ocean by 10:30 P.M. the night before (experts made the guess based on interviewing the crewmembers onboard). As a result, the first version of the search pattern was off by a significant margin. Fortunately, the Coast Guard (after engaging in a little detective work) revised their first estimate and generated a new set of search patterns. By the afternoon, the search helicopter—ironically, deviating from the computer-generated search pattern—spotted Aldridge alive in the water and retrieved him. Although the Coast Guard’s analytics weren’t wholly responsible for recovering the fisherman, the algorithm-based system is far more advanced than the centuries-old method of simply having searchers comb a grid. Big Data: it doesn’t just make systems more efficient, but potentially saves lives, as well. Image: Iakov Kalinin/Shutterstock.com How the Coast Guard Uses Analytics to Search for Those Lost at Sea
[caption id="attachment_15755" align="aligncenter" width="500"] It's a big ocean out there.[/caption] In the early-morning hours of July 24, a Montauk fisherman named John Aldridge fell off a lobster boat in the Atlantic Ocean 40 miles off Long Island, the victim of an onboard mishap with a hook. The boat motored on, its crewmembers unaware for several hours that one of their own had tumbled overboard. As Aldridge floated in the water, wondering whether he would survive the day, his shipmates radioed the U.S. Coast Guard to ask for assistance. According to The New York Times, which documented the rescue operation, the Coast Guard deployed sizable resources to find him—including a sophisticated set of data-analytics tools. Those tools include a simulation that incorporates various data-points such as a person’s weight and height, ocean temperature, and the weather to plot out the chances of survival after plunging into the water; the simulation told the Coast Guard that Aldridge would last a maximum of 19 hours before a combination of hypothermia and exhaustion sent him under. The Coast Guard also relies on software to generate a search grid. “At its heart is a Monte Carlo-style simulator that can generate, in just a few minutes, as many as 10,000 points to represent how far and in what direction a ‘search object’ might have drifted,” is how the Times describes it. “Operators input a variety of data, from the last known location of a lost mariner to the ocean currents and wind direction.” That software’s been in place for seven years, although the Coast Guard has apparently relied on simulations since the 1970s. Algorithms allow the agency to radically reduce a search area, although the output is only as good as the input; Aldridge fell in the water at 3:30 A.M., but the system initially had him in the ocean by 10:30 P.M. the night before (experts made the guess based on interviewing the crewmembers onboard). As a result, the first version of the search pattern was off by a significant margin. Fortunately, the Coast Guard (after engaging in a little detective work) revised their first estimate and generated a new set of search patterns. By the afternoon, the search helicopter—ironically, deviating from the computer-generated search pattern—spotted Aldridge alive in the water and retrieved him. Although the Coast Guard’s analytics weren’t wholly responsible for recovering the fisherman, the algorithm-based system is far more advanced than the centuries-old method of simply having searchers comb a grid. Big Data: it doesn’t just make systems more efficient, but potentially saves lives, as well. Image: Iakov Kalinin/Shutterstock.com
It's a big ocean out there.[/caption] In the early-morning hours of July 24, a Montauk fisherman named John Aldridge fell off a lobster boat in the Atlantic Ocean 40 miles off Long Island, the victim of an onboard mishap with a hook. The boat motored on, its crewmembers unaware for several hours that one of their own had tumbled overboard. As Aldridge floated in the water, wondering whether he would survive the day, his shipmates radioed the U.S. Coast Guard to ask for assistance. According to The New York Times, which documented the rescue operation, the Coast Guard deployed sizable resources to find him—including a sophisticated set of data-analytics tools. Those tools include a simulation that incorporates various data-points such as a person’s weight and height, ocean temperature, and the weather to plot out the chances of survival after plunging into the water; the simulation told the Coast Guard that Aldridge would last a maximum of 19 hours before a combination of hypothermia and exhaustion sent him under. The Coast Guard also relies on software to generate a search grid. “At its heart is a Monte Carlo-style simulator that can generate, in just a few minutes, as many as 10,000 points to represent how far and in what direction a ‘search object’ might have drifted,” is how the Times describes it. “Operators input a variety of data, from the last known location of a lost mariner to the ocean currents and wind direction.” That software’s been in place for seven years, although the Coast Guard has apparently relied on simulations since the 1970s. Algorithms allow the agency to radically reduce a search area, although the output is only as good as the input; Aldridge fell in the water at 3:30 A.M., but the system initially had him in the ocean by 10:30 P.M. the night before (experts made the guess based on interviewing the crewmembers onboard). As a result, the first version of the search pattern was off by a significant margin. Fortunately, the Coast Guard (after engaging in a little detective work) revised their first estimate and generated a new set of search patterns. By the afternoon, the search helicopter—ironically, deviating from the computer-generated search pattern—spotted Aldridge alive in the water and retrieved him. Although the Coast Guard’s analytics weren’t wholly responsible for recovering the fisherman, the algorithm-based system is far more advanced than the centuries-old method of simply having searchers comb a grid. Big Data: it doesn’t just make systems more efficient, but potentially saves lives, as well. Image: Iakov Kalinin/Shutterstock.com 